See how Conor Grogan's innovation can use AI to protect smart contracts. Learn how his use of ChatGPT-4 highlighted Ethereum smart contracts' shortcomings and why this is important for practical applications.
Director of Coinbase Conor Grogan recently achieved a huge advancement in smart contracts. He disclosed on social media that he had added a live Ethereum smart contract to ChatGPT-4, an OpenAI-trained AI language model. Unexpectedly, the AI uncovered security flaws in the contract and demonstrated how to exploit them right away.
Grogan continued by stating that the contract has already been abused by hackers in 2018. This emphasizes how crucial it is to subject smart contracts to more thorough testing and validation, especially when those contracts have practical implications. Smart contract faults and vulnerabilities can be found and fixed via AI analysis before hostile parties take use of them.
Grogan attempted to process Euler's smart contract using ChatGPT-4 in addition to testing the Ethereum smart contract. He discovered that the contract was too lengthy for the AI to process, though. Despite this setback, Grogan is confident that AI will be a key factor in making smart contracts more secure and straightforward to create.
Salus Security, a security blockchain business, examined the data supplied in the tweet and found that several of the vulnerabilities detected by ChatGPT were false positives or erroneous. However, it was later established that ChatGPT 3.5, not 4.0 as was previously claimed, was used.
It was noted that the AI didn't really help with auditing tasks despite the potential hazards mentioned by ChatGPT. Consequently, rather than actual analysis, the results resembled vulnerability enumeration.
It's crucial to remember, though, that AI has the potential to significantly enhance smart contract security in the future. Conor recognized this, saying that ultimately AI will make creating smart contracts safer and simpler.
In conclusion, while the idea of using AI for smart contract auditing is exciting, it is obvious that more work needs to be done to increase the technology's efficiency and precision.